@pmddomingos Research resulting in the 1999 Nobel Prize—awarded ”for elucidating the quantum structure of electroweak interactions"—used Schoonschip (developed by Veltman in the 1960s), which was the cornerstone of the comprehensive work of testing different ideas to renormalize the theory…
@oldfriend99@leecronin As AI trains on richer behavioral and contextual data, surely it can arrive at something at least “approximating” human memories, which involve subconscious processing of historical data, lived experiences, and subtle pattern recognition?
I found this thread very interesting and most enlightening! In particular, the observation that, in different bases, the transformers fail on the *same* numbers…
1/ The Collatz conjecture has humbled mathematicians for near 90 years. Paul Erdös said "mathematics is not yet ready for such problems." 🧵
So we handed it to transformers. What they learned stunned us!
New research from Axiom's math discovery team, blog live this morning.
@littmath Ken Ono recently joined Axiom (https://t.co/cqVSnwrcgC), which piqued my interest—and the *coupling* of AI systems to Lean also looks promising, e.g. “Building the Mathematical Library of the Future” (https://t.co/SVg06jg3kj) and https://t.co/OM5qfYZJ3s
@skdh Amazing things are definitely happening! For example, Ken Ono joined Axiom (https://t.co/H2BNwoKmvW)—and the coupling of AI systems to Lean looks promising, such as “Building the Mathematical Library of the Future” (https://t.co/SVg06jg3kj) and https://t.co/OM5qfYZJ3s